MICROFABRICATION Alessandro MAPELLI - DT Group Meeting, 22 June 2017 - CERN Indico
←
→
Page content transcription
If your browser does not render page correctly, please read the page content below
Alessandro MAPELLI DT Group Meeting, 22 June 2017 MICROFABRICATION Anastasia Jacopo Timothée Georgi Ranit Luca Clémentine Ludovic BERDNIKOVA BRONUZZI FREI GORINE MONGA MÜLLER LIPP SEREX
MICROFABRICATION OF ON-DETECTOR COOLING SYSTEMS Fabrication of structures with features below the mm. Same techniques as used for microelectronics and silicon detectors. Novel high performance compact systems based on microfluidics. ical d on Plastic Opt til la tio n Pa rti cl e Detectors Base Scin (2011) THÈSE N 5033 O 2011 2 SEPTEMBRE PRÉSENTÉE LE HN IQU ES DE L'INGÉNIEUR IENCES ET TEC g À LA FACULTÉ SC i n ES 4 l TÈM o RO SYS o DE MIC c LABORATOIRE CTRONIQUE el ES ET MICROÉLE an n MICROSYSTÈM Silicon microch CTO RA L EN PROGRAMME DO ANNE RALE DE LAUS CHNIQUE FÉDÉ ÉCOLE POLYTE SCIENCES DE DOCTEUR ÈS ON DU GRADE POUR L'OBTENTI w ell, A. Klug e, A. Map elli, M. Morel, J. Noel, guin, R. Dumps, K. Ho J. Buytaert, A. Catinaccio, J. Da PAR L), G. Ro m ag no li, J. Thome (EPFL) LI ssle, P. Petagna, Ph. Re ELue MAPN na ud (E PF AlessandroG. Nuclear Physics B (Proc. Suppl.) 215 (2011) 349–352 www.elsevier.com/locate/npbps position du jury : acceptée sur pro Prof. M. A. Ionesc Prof. Ph. Renaud u, président du jury , Dr P. L. G. Grafstr öm, directeurs de , rapporteur thèse Low material budget microfabricated cooling devices for particle Prof. A. Bay Dr B. Gorini, rap Dr M. Haguenaue por teu r r, rapporteur detectors and front-end electronics A. Mapelliab ∗ , A. Catinaccioa , J. Daguina , H. van Lintelb , G.Nuesslec , P. Petagnaa , P. Renaudb , a Physics Department, CERN, Geneva, Switzerland b Laboratoire de Microsystèmes, Ecole Polytechnique Fédérale de Lausanne, Switzerland Suisse 2011 c CP3, Université catholique de Louvain, Louvain-la-Neuve, Belgique Novel cooling systems with very low material budget are being fabricated and studied. They consist of silicon wafers in which microchannels are etched and closed by bonding another wafer. This cooling option is being considered for future HEP detectors of the sLHC and linear colliders. It is currently under investigation as an DT Group Meeting, option22 forJune 2017 of the NA62 Gigatracker silicon pixel detector and its front-end the cooling Alessandro electronicsMapelli | 2 where the microfabricated cooling plate would stand directly in the beam. In this particular case, microchannel cooling
WHY MICROFLUIDIC ON-DETECTOR COOLING SYSTEMS ? d Integrated Micro Micro-channel channel Heat Sinks : q P (∆T fluid-sensor) TFM = Tj,max Tf,in (power density) Rtot ower approach TFM conventional D.B. Tuckerman and R.F.W. Pease, IEEE Elec. Dev. Letters, Tuckerman D.B. and Pease Vol. 2, R.F.W., 5, 1981 IEEE Elec. Dev. Letters, Vol. 2, 5, 1981. 20 ▸ No CTE mismatch Monday, October 11, 2010 © 2009 IBM Corporation integrated ▸ Low material budget 12 ▸ Active/distributed cooling ▸ Radiation resistance microchannels 5-8 ▸ Great integration potential liquid ▸ Thermal Figure of Merit 3 bi-phase DT Group Meeting, 22 June 2017 Alessandro Mapelli | 3
HOW DO WE MANUFACTURE SILICON MICROCHANNELS ? ▸ Collaborative effort between experiments (ALICE, LHCb and NA62), support groups (DT and ESE), and external partners (CSEM and EPFL). photolithography layout 8” wafer processed at CEA-LETI µchannels etching & Wafer Bonding Wafer Thinning cross-section of the cooling plates with embedded microchannels Silicon Etching DT Group Meeting, 22 June 2017 Alessandro Mapelli | 4
NA62 GTK - PIONEERING MICROCHANNELS ! ▸ 2009 - concept of micro-cooling for the GTK presented Pπ to NA62: https://indico.cern.ch/event/58370/ PK 2 3 1 K_ K_ K_ ▸ 2014 - First GTK in the experiment GT GT GT P ▸ 2016 - Data taking with 3 GTK detectors P ▸ 2017 - Assembly of 6 GTK modules for 2018 13 m 9m GTK_3 9m GTK_2 13m GTK_1 Jordan DEGRANGE Jerome NOEL Diego ALVAREZ FEITO contributed to the design of the module The beam and detector of the NA62 experiment at CERN, Journal of Instrumentation, Volume 12, May 2017, https://goo.gl/P391U3 DT Group Meeting, 22 June 2017 Alessandro Mapelli | 5
LHCb VELO UPGRADE - EVAPORATIVE CO2 IN MICROCHANNELS Feature Current VELO Upgraded VELO Sensors # of modules ▸ Primary heat source due to the VeloPix chips (~1 W/cm2) R & strips 42 0.22 m2 Pixels 52 0.12 m2 ▸ Detector Active area Technology Sensors must be kept at -20ºC ⇠172k strips electron collecting ⇠41M pixels electron collecting Max fluence ▸ Evaporative CO2 circulates in microchannels routed directly under the VeloPix chips. 300 µm thick 3.9 ⇥ 1014 200 µm thick 8 ⇥ 1015 ▸ MeV · neq /cm 2 MeV · neq /cm 2 HV tolerance ASIC Readout rate Total power consumption > 2kW. 500 V 1 MHz 1000 V 40 MHz Total data rate ▸ 52 modules in secondary vacuum separated by primary beam vacuum by a 0.25 Total Power consumption ⇠150 Gb/s ⇠ 1 kW 1.2 Tb/s 2.2-2.3 kW mm thick Al foil. Table 1: An outline of the major di↵erences between the current VELO and its upgrade ▸ 24 cooling plates delivered by CEA-Leti. 20 more will be delivered within 2 months. ▸ Void-less soldering of connector to cooling plates procedure under development. Figure 2: (top) Schematic of one half of the VELO upgrade detector. Twenty- six modules are aligned along the beam direction. (bottom) Two modules in the closed position. The modules are in di↵erent z-positions along the beam line to minmise gaps in the acceptance and there is a slight overlap between the two halves which aids detector alignment. c Copyright NIKHEF: reproduced with permission rection, with the full length of the detector being approximately 1m. A module is made of two half-disc sensors with R- and -measuring geometry. 3. VELO Upgrade Figure 3: 3D Impact Parameter resolution vs particle inverse transverse mo- The Vertex Locator upgrade [3] is a significant redesign from mentum. Made using LHCb software simulation. A significant resolution im- the original detector. The major changes in new detector and its provement of the upgraded detector (red) can be seen over the current VELO (black). The upgrade operates closer to the beam line and has a lower material predecessor are as follows: budget - two major contributions to the improved resolution. c Copyright CERN: reproduced with permission • The detector will change from a silicon strip detector to a pixel detector. budget with thinner sensors and and thinner aluminium foil • The detector will be closer to the beam in its closed posi- housing each detector half and (ii) placing the sensors closer tion at 5.1 mm from 8.2 mm to the beam. The closer placement of the detector modules to • The upgraded VELO will use a new VeloPix ASIC [6] that the beam and the higher luminosity for LHC Run III comes with can be read out at 40 MHz (up from 1.1 MHz) and a band- some costs - the detector must be able to handle higher radiation width of up to 20.4 Gb/s. doses; there is a higher hit occupancy due to increased particle flux. This, in turn, results in higher data rates from the detector • The current detector modules are cooled with CO2 pass- and greater power consumption in the front-end ASICs (how- ing through a series of cooling blocks attached to the base ever, the dominant increase in data rate is due to the removal of of the module substrate. The upgraded modules will also the L0 hardware trigger). Table 1 lists some of the major di↵er- use CO2 as a coolant, but it will pass through micro- ences between the current VELO detector and the upgrade. channels in a silicon substrate, directly beneath the major heat sources (VeloPix ASICs and other chips). 3.1. Modules The upgraded VELO (shown in figure 2) will have more ro- A mechanical module concept is shown in figure 4. The cur- bust track reconstruction performance compared to a strip de- rent design of the module consists of a carbon-fibre structure tector and an overall improved resolution (figure 3). This im- supporting a silicon microchannel substrate. Stress-relieved proved performance is achieved by (i) lowering the material CO2 cooling pipes route the CO2 to and from the cooling con- 2 SILICON COOLING PLATE MOCKUP WITH CONNECTOR AND CAPILLARIES DT Group Meeting, 22 June 2017 Alessandro Mapelli | 6
SILICON THERMAL MOCKUPS heater microchannel Resistive Temperature Detector microfluidic inlet microfluidic outlet inlet capillary outlet manifold soldering pad soldering pad ▸ Timothée FREI, PhD EPFL, 2021 ▸ processing at CSEM on 6” wafers ▸ Désirée HELLENSCHMIDT, PhD ▸ post-processing at EPFL of 6” wafers from FBK ▸ Ranit MONGA, Intern, 2017 ▸ processing at EPFK on 4” wafers DT Group Meeting, 22 June 2017 Alessandro Mapelli | 7
INTEGRATING MICROCHANNELS ON PIXEL DETECTORS ▸ Development of large-area monolithic detectors ▸ Embed microchannels on the backside of CMOS wafers “LOW TEMPERATURE” WAFER BONDING BURIED CHANNELS CEA-Leti, EPFL, G-Ray EPFL, FBK COOLING PLATE EMBEDDED MICROCHANNELS COOLING PLATE EMBEDDED MICROCHANNELS trenches - anisotropic etch trenches - anisotropic etch channels - anisotropic etch channels - anisotropic etch arch - Zurich channels - isotropic etch channels - isotropic etch quid Cooled Integrated Micro Micro-channel channel Heat Sinks wafer bonding wafer bonding Figure of Merit: q COP P trench filling trench filling Heat Flux: Tj,max Tf,in q Achip Rtot Pumping Power P V p D.B. Tuckerman Tuckerman D.B. andand R.F.W. Pease Pease, R.F.W., IEEE IEEE Elec. Elec. Dev.Dev. Letters, Vol. 2, 5,Letters, 1981. Vol. 2, 5, 1981 Jacopo BRONUZZI, PhD EPFL, 2018 Monday, October 11, 2010 © 2009 IBM Corporation Clémentine LIPP, MSc EPFL, 2017 DT Group Meeting, 22 June 2017 Alessandro Mapelli | 8
jectives and Aims INTERCONNECTIVITY - DAISY CHAINING MICROFLUIDICS esent work was carried out in the framework of the development of a silicon ▸ NA62, LHCb Proposal by Miranda VAN STENIS hannel-based cooling solution for the ALICE ITS. In order to illustrate the objectives 1/16 SS pipe LASER soldered to KOVAR ms of the work, an overview of a stave (i.e. the basic subcomponent ofMasterthe ▸ limited surfaces Thesis ITS) Development of an in plane microfluidic interconnect (CTE 6 ppm/K) Tiago Morais Cu 21mm for the thermal management of particle detectors plating ng such a cooling solution is shown in Figure NA62 5. Basically, it is composed Aof 1-kilosilicon ▸ access from the side 2 weight was used to press the syringe piston which has a surface of 550 mm . The SnPb pressure at the syringe is calculated as follows: used to track the particles 0.1 mm Foil resulting from collisions) glued to a silicon frame containing 1 ∗ 9.81 ▸ alarger surfaces Au 100 nm −2 Ni 350 nm 3.6mm = = = 550 = 1.78 ∗ 10 = 178 2 hannels. A coolant Ti 200isnmsupplied to the microchannels through a connector, and The test is considered successful if water starts filling the cup and no drop of water is observed are▸ interconnections illepaM.A Silicon on or around the sample after 5 minutes of water circulation. fibre structure stiffens the assembly (furtherLHCb details on the layout of the Two (CTE 3 ppm/K) stavedifferent samples were tested with this procedure, they both circulate water. All the remaining samples were tested previously without using the outlet clamping. When pressing n section 3). the syringe, drops of water flowed from the free outlet. In addition, some samples were cut near the connector, as shown in Figure 5.2 a), to allow imaging of the interior of the interconnect. igure 5 : Exploded-view of a stave for the ALICE ITS, including a silicon microchannel-based cooling system. Figure 5.2 a) Test sample with dicing path in red and cross-section orientation, b) Cross- DT Group Meeting, 22 June 2017 Alessandro sectional SEM image of the sample showing the two layers (Borofloat Mapelli 33 and silicon) and|the9 interconnect
SUMMARY & OUTLOOK ▸ Inter-disciplinary efforts are required to develop, fabricate and operate microsystems and microfluidic devices. ▸ DT is the perfect environment for such developments. ▸ The NA62 GTK modules and LHCb Velo cooling plates involve expertise from every section of the group. ▸ NA62 GTK - fabrication of cooling plates, assembly of modules and installation in the experiment. ▸ LHCb Velo - fabrication of the cooling plates and development of the soldering procedure for the connector. ▸ Silicon thermal mockups design and fabrication guidelines. ▸ Better integration of services in silicon sensors by embedding microchannels. ▸ Microfluidic interconnections to daisy chain cooling plates for staves ▸ Microfluidic closed-loop heat pipes for space applications. DT Group Meeting, 22 June 2017 Alessandro Mapelli | 10
You can also read